Processing model of multi-scale geospatial data based on genetic algorithms
Hongyan Deng, Fang Wu, Qian Zhao, Dongmei Dong · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2007
It is one of the most important and far-reaching problems about multi-scale processing and representation of geospatial data in geographic information science. Processing model of multi-scale geospatial data is the key to the problem. After deeply analysing principles of Genetic Algorithms, a processing model of multi-scale geospatial data based on Genetic Algorithms is proposed: 1.determining coding, this model used restricted coding method combined with existing models; 2. making fitness function: the geometric feature of points cluster and the number of points in line are leading guidelines of fitness function; 3. ascertaining local optimization strategy: it takes contrast of points cluster and precision of points in line as the secondary factors, in order to achieve high optimization efficiency. Experiments have demonstrated that the model does well in terms of preserving geometric feature of geospatial data.